Behavioural modelling of road users: current research and future needs

Abstract

Despite the considerable work done on travel behaviour in general and on driving behaviour in particular, it is argued that most of the behavioural models still lack a cognitive explanatory mechanism of the individual’s choice process. The paper presents a survey of recent important research in this area from European and North American perspectives in order to identify issues that should be studied more closely as a base for a new research agenda. It was found that since the human cognitive mechanism of travel decision‐making is universal, idiosyncratic situations, cultural and societal norms can affect the individual’s perception of constraints that will consequently affect the whole decision‐making process. An extension of the Decision Field Theory is proposed as a framework for a new research agenda, which will include the effects of travel situations (e.g. timing, dynamics and type) as well as of cultural habits and societal norms. This theory is aimed at understanding the motivational and cognitive mechanisms that guide a deliberation process involved in making travel decisions under uncertainty.

Notes

Correspondence Address: Eliahu Stern, Department of Geography & Environmental Development, Ben‐Gurion University of the Negev, Beer Sheva 84105, Israel. Email: [email protected]

Technology has long been useful in travel behaviour research in the form of simulators to test hypotheses about driving under laboratory conditions (Brookhuis et al., 1996). However, in more recent years and in the near future, technology has played, and is playing, a more important role in terms of on‐board equipment to correct for aberrant driving behaviour. There are major technological changes either in place or soon to be introduced that offer automatic remedies for many driving behaviour problems. These include antilock brakes, breathalyser ignition locks, traction and other electronic stability controls, adaptive cruise control, electric steering, rear parking warning systems, mechanisms to turn off cellphones and other noise or visual distractions when driving conditions become difficult, self‐parking mechanisms, global positioning satellite (GPS) guidance controls, on‐board intelligent transport systems (ITS) systems about upcoming congestion, airbag deactivation for small children under specific crash conditions, etc. Another example is eye‐tracking monitors to measure visual alertness; Ford and Volvo are engaged in research to develop on‐board technology to monitor eye movements and to transmit warnings (lights, vibration, buzzers) to drivers either falling asleep or at risk of doing so. This technology is expected to be in production vehicles by the end of the decade. As for reducing congestion, autonomous (i.e. driverless) vehicles offer promise (a USDOT programme from the 1990s was a technological success, but nevertheless was cancelled in 1998), as does platooning (especially of commercial trucks, reducing labour costs by requiring only a lead vehicle driver). Two important questions arise with respect to technology. Will drivers accept it? The major problem with some of these innovations might be less the technology but consumer acceptance; drivers are likely to welcome some technological advances much more willingly than are others. For example, adaptive cruise control might gain ground much faster than fully autonomous vehicles. Are drivers willing to pay for new technology? The costs of the equipment will vary widely and there might be little correlation between cost and safety benefits. Also, some drivers value safety more than others, while other drivers assess these technological advances primarily as convenience items that have little impact on driving behaviour.

Traffic psychologists have given most attention to driving behaviour that impacts upon safety (Sivak, 2002). Major issues discussed include drowsiness or falling asleep at the wheel; the problems associated with very young or elderly drivers (note, however, that motor vehicle deaths per 100 000 are lowest in the 55–59‐year age group [12], while even the 80–84‐year age group has a lower rate [23] than the 16–19‐[31] and 20–24‐year age groups [28], NHTSA data); attitudes towards speed, tailgating and changing lanes; road rage; and alcohol and drug abuse. With respect to the role of drowsiness, sleep‐related accidents of truck drivers increase with the length of the trip (in excess of guidelines or laws) and with speed. However, economic and social factors might also be influential. For example, there is evidence that higher wages and occupational benefits reduce this type of crash, probably by relieving pressures of time. Also, accident rates reflect not only differences in vehicle size and safety features, but in how they are driven‐in other words, social psychological connections between types of drivers, the vehicles to which they are attracted and the ways in which they are driven. According to US Federal data, the rate of all driver deaths (for 1994–97 model years) was 89 per million registered vehicle‐years. However, the rates ranged from 20 for the Infiniti J30 and 37 for the Toyota Camry to 209 for the GeoMetro and 308 for the Chevy Camaro. Referring to more recent data (2000–02), the variation in injury rates (vehicle model average index = 100) ranged from 44 for the Buick Le Sabre to 239 for the Suzuki Esteem, with 56 for the safest sports sedan, the Saab 9‐3. The point is that vehicle choice, often primarily a psychological decision, may affect risk and safety, and certainly driving behaviour.

A real‐world example might illustrate the point. A tram shuttle operates for an 8‐mile trip between the University of Southern California’s two campuses in Los Angeles. The outbound trip from the main campus has an intermediate stop downtown at the rail station (Union Station); the inbound trip is non‐stop. The service employs five drivers. They all choose the same inbound route via two freeways with only minor deviations (alternative exits); this is consistent with the utility‐maximizing (time‐minimizing) route choice model. The outbound trip is a different story. In addition to two or three freeways, there is considerable network redundancy on surface streets. Each driver habitually chooses a different route, and route choice appears to be insensitive to the time of travel (e.g. peak or off‐peak). Presumably, their mental maps vary and the choice of different routes is consistent with fuzzy individual preference‐based theory.

Additional information

Notes on contributors

ELIAHU STERN

Correspondence Address: Eliahu Stern, Department of Geography & Environmental Development, Ben‐Gurion University of the Negev, Beer Sheva 84105, Israel. Email: [email protected]